I browsed this section of the forum but wasn't able to find anything related to this topic. I'm currently evaluating your product and one of the features I've been asked to examine is if the engine is capable of recognizing handwritten text. I've already run some tests and my conclusion is that the the letter is only recognized if it is written as if it was typed. Different people have different style of writing the letter or digits, for example one could add a hyphen(minus) in the middle of the digit 7 and this won't be recognized as a 7, so my question is if it's possible to build some kind of a template, as you do with the dictionaries, against which we are going to compare the letter read. The term self-teaching-algorithm was used somewhere in the previous posts, which I'm not familiar with, but I wonder if it could be applied in this case - whenever we read a scanned document with handwritten text we teach our system (update the template) if a match with a given level of certainty wasn't found in our template. I look forward to hearing from you.
Who is online
Users browsing this forum: No registered users and 1 guest